dailysudoku.com :: View topic - Free Press Mar 5, 2010

dailysudoku.com
Discussion of Daily Sudoku puzzles

FAQ

Memberlist

Usergroups

Profile

Free Press Mar 5, 2010

dailysudoku.com Forum Index -> Other puzzles

View previous topic :: View next topic

Author

Message

keith

Joined: 19 Sep 2005
Posts: 3355
Location: near Detroit, Michigan, USA

Posted: Fri Mar 05, 2010 6:32 pm Post subject: Free Press Mar 5, 2010

A long slog.

Code:

Puzzle: FP030510
+-------+-------+-------+
| 1 . . | . . . | 7 . . |
| 2 . 3 | . . 4 | . 6 1 |
| . 6 . | . . . | . 2 . |
+-------+-------+-------+
| 5 . . | . 7 1 | . 3 . |
| . . . | . . . | . . . |
| . 3 . | 5 4 . | . . 8 |
+-------+-------+-------+
| . 1 . | . . . | . 4 . |
| 6 4 . | 3 . . | 9 . 7 |
| . . 8 | . . . | 2 . 5 |
+-------+-------+-------+

Keith

tlanglet

Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills

Posted: Fri Mar 05, 2010 10:11 pm Post subject:

Four steps.

Quote:

xy-wing 4-89 with vertex in r1c3 and pseudocell 89 in box3 deletes 8 in r2c2 and r3c7,
skyscraper 7 in r39c6 deletes 7 in r2c2, which sets up a
w-wing 59 in r1c8 & r2c2 with strong link on 9 in r6c18 with transport: (9)r6c1 - r7c1 = r9c2 that deletes 5 in r1c2,
coloring on 4 completes the puzzle.

Ted

arkietech

Joined: 31 Jul 2008
Posts: 1834
Location: Northwest Arkansas USA

Posted: Sat Mar 06, 2010 12:08 am Post subject:

Two steps with another "Dan's step with no name" Very Happy

Code:

skyscraper 7
*-----------------------------------------------------------------------------*
| 1 589 49 | 2689 235689 3689 | 7 59 349 |
| 2 589 3 | 7 589 4 | 58 6 1 |
| 48 6 7 | 1 3589 389 | 3458 2 349 |
|-------------------------+-------------------------+-------------------------|
| 5 289 469 | 689 7 1 | 46 3 2469 |
| 48 2789 1469 | 689 3689 3689 | 1456 579 2469 |
| 79 3 16 | 5 4 2 | 16 79 8 |
|-------------------------+-------------------------+-------------------------|
| 79 1 5 | 289 289 789 | 36 4 36 |
| 6 4 2 | 3 1 5 | 9 8 7 |
| 3 79 8 | 4 69 679 | 2 1 5 |
*-----------------------------------------------------------------------------*
(3=4)r1c9 = r1c3 - (4=8)r3c1 - (8=59)r1c238 = (3)r1c9; r1c9=3

Most likely illegal but lots of fun. Rolling Eyes

Marty R.

Joined: 12 Feb 2006
Posts: 5770
Location: Rochester, NY, USA

Posted: Sat Mar 06, 2010 12:19 am Post subject:

A long slog indeed, but an enjoyable BBDB. A couple of ERs, Type 4 UR, X-Wing on 2, then my key move followed by Remote Pairs (79), W-Wing (38) and Type 1 UR (28). Maybe someone can name my key move which I referenced. I've always called them XY-Chains which purists won't like because they involve more than bivalue cells. But it works similarly in that you start out like an XY-Chain and establish pincers. It can easily be proved that if r6c8=7, then r1c3=9, thus the pincers take out the 9 from r1c8. So, what is it?

storm_norm

Joined: 18 Oct 2007
Posts: 1741

Posted: Sat Mar 06, 2010 3:04 am Post subject:

the three step journey is
skyscraper on 7
to this grouped two-string kite on 9

Code:

+--------------------+--------------------+------------------+
| 1 589 4(9) | 2689 235689 3689 | 7 5-9 349 |
| 2 589 3 | 7 589 4 | 58 6 1 |
| 48 6 7 | 1 3589 389 | 3458 2 349 |
+--------------------+--------------------+------------------+
| 5 289 46(9) | 689 7 1 | 46 3 2469 |
| 48 2789 146(9) | 689 3689 3689 | 1456 579 2469 |
| 7(9) 3 16 | 5 4 2 | 16 7(9) 8 |
+--------------------+--------------------+------------------+
| 79 1 5 | 289 289 789 | 36 4 36 |
| 6 4 2 | 3 1 5 | 9 8 7 |
| 3 79 8 | 4 69 679 | 2 1 5 |
+--------------------+--------------------+------------------+

then skyscraper on 4 finishes.

tlanglet

Joined: 17 Oct 2007
Posts: 2468
Location: Northern California Foothills

Posted: Sat Mar 06, 2010 12:04 pm Post subject:

arkietech wrote:

Two steps with another "Dan's step with no name" Very Happy

Code:

Most likely illegal but lots of fun. Rolling Eyes

Dan, I am having trouble understanding this step, logically and notationally. I have a guess about the logic, but the notation does not use alternating strong and weak links. Some prose would be very helpful.

Thanks,
Ted

arkietech

Joined: 31 Jul 2008
Posts: 1834
Location: Northwest Arkansas USA

Posted: Sat Mar 06, 2010 1:14 pm Post subject:

tlanglet wrote:

You are right the coding is in error. It was very clear to me last night. Embarassed

I will try to recreate my logic as I wake up.
update:

Code:

*-----------------------------------------------------------------------------*
| 1 589 49 | 2689 235689 3689 | 7 59 349 |
| 2 589 3 | 7 589 4 | 58 6 1 |
| 48 6 7 | 1 3589 389 | 3458 2 349 |
|-------------------------+-------------------------+-------------------------|
| 5 289 469 | 689 7 1 | 46 3 2469 |
| 48 2789 1469 | 689 3689 3689 | 1456 579 2469 |
| 79 3 16 | 5 4 2 | 16 79 8 |
|-------------------------+-------------------------+-------------------------|
| 79 1 5 | 289 289 789 | 36 4 36 |
| 6 4 2 | 3 1 5 | 9 8 7 |
| 3 79 8 | 4 69 679 | 2 1 5 |
*-----------------------------------------------------------------------------*
if r1c2<>8 r1c2 and r1c8 form a 59 pair and r1c3=4 => r1c9=3
(3=4)r1c9 = r1c3 - (4=8)r3c1 - [(8)r1c2 = (3)r1c9]; r1c9=3

Brackets indicates what I am trying to code and don't know how.

storm_norm

Joined: 18 Oct 2007
Posts: 1741

Posted: Sat Mar 06, 2010 4:10 pm Post subject:

arkietech wrote:

tlanglet wrote:

You are right the coding is in error. It was very clear to me last night. Embarassed

I will try to recreate my logic as I wake up.
update:

Code:

*-----------------------------------------------------------------------------*
| 1 589 49 | 2689 235689 3689 | 7 59 349 |
| 2 589 3 | 7 589 4 | 58 6 1 |
| 48 6 7 | 1 3589 389 | 3458 2 349 |
|-------------------------+-------------------------+-------------------------|
| 5 289 469 | 689 7 1 | 46 3 2469 |
| 48 2789 1469 | 689 3689 3689 | 1456 579 2469 |
| 79 3 16 | 5 4 2 | 16 79 8 |
|-------------------------+-------------------------+-------------------------|
| 79 1 5 | 289 289 789 | 36 4 36 |
| 6 4 2 | 3 1 5 | 9 8 7 |
| 3 79 8 | 4 69 679 | 2 1 5 |
*-----------------------------------------------------------------------------*
if r1c2<8> r1c9=3
(3=4)r1c9 = r1c3 - (4=8)r3c1 - [(8)r1c2 = (3)r1c9]; r1c9=3

Brackets indicates what I am trying to code and don't know how.

hey Dan,
I'll try and explain this. what you have is a ALS on cells r1c2389
you noticed that if the 8 in r1c2 and the 3 in r1c9 are both false then you would have an illegal state of the grid.
you would have 59 in two cells and 49 in two cells in row 1. this can't happen.

so you can write this as: ALS[(3)r1c9 = (8)r1c2]r1c2389
one of the 3 or the 8 must be true to avoid the illegal state of the grid.

secondly,
your bivalue cell in r3c1 {4,8} only completes a loop with the ALS.

in a ALS like the one you found, all of the candidates in each cell are strongly inferenced. which means all of the similar candidates that see each other are weakly linked.

well guess what? this happens even with the bi-value cell {4,8} in r3c1. the 4 is weakly linked to the ALS.

Code:

- ALS[(3)r1c9 = (8)r1c2]r1c2389 - (8=4)r3c1 -
| |
|-------------------------------------------|

the weakly linked 4 completes the loop back to the front
because the 4's are part of the candidates in the ALS in row 1

this is tricky to keep track of, but in a loop, all of the weakly linked candidates now become strongly linked
which means all of those weakly linked
9's
5's
4's
all make elminations in row 1.
and the weakly linked 8's in box 1.

creates a nice little picture.

unfortunately, this loop does not place a 3 in r1c9 immediately.
your only mistake was the first part of the chain under your grid.

(3=4)r1c9 = r1c3 - (4=8)r3c1 - [(8)r1c2 = (3)r1c9];

the red colored strong link should be a weak link and isn't really necessary because the 3 and the 4 reside in the same cell giving them a weak inference which completes the loop. And this would eliminate the 9 in r1c9 not place the 3.

arkietech

Joined: 31 Jul 2008
Posts: 1834
Location: Northwest Arkansas USA

Posted: Sat Mar 06, 2010 4:36 pm Post subject:

storm_norm wrote:

unfortunately, this loop does not place a 3 in r1c9 immediately.

Thanks Norm for correcting my magic 'look-ahead' loop. I will study this carefully.

keith

Joined: 19 Sep 2005
Posts: 3355
Location: near Detroit, Michigan, USA

Posted: Sat Mar 06, 2010 6:58 pm Post subject:

Here's another of those M-things. After basics:

Code:

+----------------------+----------------------+----------------------+
| 1 589 49 |268-9 23568-9 368-9 | 7 59a 349 |
| 2 5789b 3 | 789c 589c 4 | 58 6 1 |
| 48 6 479 | 1 3589 3789 | 3458 2 349 |
+----------------------+----------------------+----------------------+
| 5 289 469 | 689 7 1 | 46 3 2469 |
| 48 2789 14679 | 689 3689 3689 | 1456 579 2469 |
| 79 3 16 | 5 4 2 | 16 79 8 |
+----------------------+----------------------+----------------------+
| 79 1 5 | 2789 289 789 | 36 4 36 |
| 6 4 2 | 3 1 5 | 9 8 7 |
| 3 79 8 | 4 69 679 | 2 1 5 |
+----------------------+----------------------+----------------------+

5 in a forces 5 in b. Not 9 in b forces 9 in one of c. Making the eliminations shown. (It does not solve the puzzle.)

Keith

Display posts from previous:

	dailysudoku.com Forum Index -> Other puzzles	All times are GMT
Page 1 of 1

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum